本講義では、教師なし学習と教師あり学習の基本概念と発展的手法について学ぶ。
教師なし学習では、確率・統計の基本的な概念、グラフィカルモデル、統計的因果探索に焦点を当て、データ間の隠れた構造や因果関係を明らかにする手法を学ぶ。
教師あり学習では、最小二乗法による線形回帰、学習と過学習の理解、カーネル法や加法モデルに焦点をあて、ブラックボックスでない非線形回帰手法を通じてモデルを理解する方法について学ぶ。
This course covers the fundamental concepts and advanced techniques in both unsupervised and supervised learning.
In the unsupervised learning section, the focus is on basic concepts in probability and statistics, graphical models, and statistical causal discovery, learning methods to reveal hidden structures and causal relationships in data.
In the supervised learning section, the focus is on linear regression through the least squares method, learning and overfitting, and kernel methods and additive models, learning methods to understandn non-black-box models through nonlinear regression techniques.